Fatty liver - steatosis - affects about one third of the population. Its prevalence is rising and seems to parallel the global increase in obesity and type-2 diabetes. The mechanisms underlying steatosis and leading to non-alcoholic fatty liver disease are poorly understood. We propose that the hepatic conversion of carbohydrates (CHO) to lipids (de novo lipogenesis, DNL) is a key factor in the accumulation of excess liver fat and the accompanying dyslipidemia and that suppressing DNL by diet will reduce liver fat and improve the metabolic profile in patients with steatosis. These hypotheses are based on studies in which we and others have established that fractional hepatic DNL can vary dramatically depending on the diet and/or health status of a subject; and, in particular, that dietary fructose is a potent lipogenic stimulus. In this proposal we will perform Clinical Research Center (CRC) based studies to compare the rates of DNL and VLDL kinetics in steatotic and matched non-steatotic controls and evaluate their relationship to lipid profiles (Aim 1). The steatotic individuals whose habitual intake of fructose and other simple sugars exceeds 15% of total energy intake will then be randomized to consume one of two low-fat diets that differ only in CHO type to determine whether diet-induced changes in DNL affect liver fat flux, content (Aim 2). We hypothesize that a diet that is rich in complex CHO will achieve greater decreases in DNL and liver fat than one that contains typical amounts of simple CHO, including fructose. This dietary intervention study includes a 6-week 25% energy restriction outpatient phase to promote moderate weight loss and improve insulin sensitivity, followed by a week weight maintenance with the last 5 days as an inpatient stay during which all of the studies performed at baseline (Aim 1) will be repeated. State-of-the-art stable isotope techniques will be used to assess hepatic DNL, apoB100 turnover, VLDL-TG fluxes, and lipolysis under fasting and fed conditions. Liver fat will be measured by proton magnetic resonance spectroscopy. These studies will allow us to evaluate the importance of DNL as a mechanism modulating liver fat content and flux, and the significance of CHO quality in dietary guidance for steatotic patients.